KarmenLoop/personal-agent-growth

[Credential-related source omitted]

Copyright (C) 2007 Free Software Foundation, Inc. <https://fsf.org/>

You should have received a copy of the GNU Affero General Public License     along with this program.  If not, see <https://www.gnu.org/licenses/>.

along with this program.  If not, see <https://www.gnu.org/licenses/>.

if any, to sign a "copyright disclaimer" for the program, if necessary. For more information on this, and how to apply and follow the GNU AGPL, see <https://www.gnu.org/licenses/>.

<https://www.gnu.org/licenses/>.

Copyright (C) 2026 KARMENLOOP Official Team Contact: info@karmenloop.com Project: https://github.com/KarmenLoop/personal-agent-growth

Contact: info@karmenloop.com Project: https://github.com/KarmenLoop/personal-agent-growth SPDX-License-Identifier: AGPL-3.0-only

Project: https://github.com/KarmenLoop/personal-agent-growth

You should have received a copy of the GNU Affero General Public License along with this program. If not, see https://www.gnu.org/licenses/.

along with this program. If not, see https://www.gnu.org/licenses/.

[![Checks](https://github.com/KarmenLoop/personal-agent-growth/actions/workflows/check.yml/badge.svg)](https://github.com/KarmenLoop/personal-agent-growth/actions/workflows/check.yml)

[![Release](https://img.shields.io/github/v/release/KarmenLoop/personal-agent-growth)](https://github.com/KarmenLoop/personal-agent-growth/releases/latest)

Download the **complete delivery bundle** and `SHA256SUMS.txt` from [v0.7.0](https://github.com/KarmenLoop/personal-agent-growth/releases/tag/v0.7.0). Extract it and follow [START-HERE.md](START-HERE.md). Python 3.11+ is needed for the util

Created and maintained by the **KARMENLOOP Official Team**, Shanghai, China. We build Koder and KARMEN BOT. [Team](TEAM.md) · [Website](https://karmenloop.com) · [GitHub](https://github.com/KarmenLoop) · [Email](mailto:info@karmenloop.com)

[![检查](https://github.com/KarmenLoop/personal-agent-growth/actions/workflows/check.yml/badge.svg)](https://github.com/KarmenLoop/personal-agent-growth/actions/workflows/check.yml)

[![发布](https://img.shields.io/github/v/release/KarmenLoop/personal-agent-growth)](https://github.com/KarmenLoop/personal-agent-growth/releases/latest)

从 [v0.7.0 发布页](https://github.com/KarmenLoop/personal-agent-growth/releases/tag/v0.7.0)下载**完整交付包**与 `SHA256SUMS.txt`，解压后按[开始使用](开始使用.md)操作。工具需要 Python 3.11+，运行时无需安装 pip 依赖。

由来自中国上海的 **KARMENLOOP Official Team** 创建与维护。团队开发 Koder 与 KARMEN BOT。[团队介绍](TEAM.zh-CN.md) · [官网](https://karmenloop.com) · [GitHub](https://github.com/KarmenLoop) · [邮箱](mailto:info@karmenloop.com)

| Contact | Details | |---|---| | Website | [karmenloop.com](https://karmenloop.com) |

|---|---| | Website | [karmenloop.com](https://karmenloop.com) | | GitHub | [KarmenLoop](https://github.com/KarmenLoop) |

| Website | [karmenloop.com](https://karmenloop.com) |

| GitHub | [KarmenLoop](https://github.com/KarmenLoop) |

| Public contact page | [Contact KARMENLOOP](https://karmenloop.com/en/contact.html) |

| 联系方式 | 地址 | |---|---| | 官方网站 | [karmenloop.com](https://karmenloop.com) |

|---|---| | 官方网站 | [karmenloop.com](https://karmenloop.com) | | GitHub | [KarmenLoop](https://github.com/KarmenLoop) |

| 官方网站 | [karmenloop.com](https://karmenloop.com) |

| GitHub | [KarmenLoop](https://github.com/KarmenLoop) |

| 公开联系页 | [联系 KARMENLOOP](https://karmenloop.com/en/contact.html) |

The scientific input came from the official [NASA GISTEMP CSV](https://data.giss.nasa.gov/gistemp/tabledata_v4/GLB.Ts+dSST.csv), retrieved October 23, 2026, with [v4 source documentation](https://data.giss.nasa.gov/gistemp/data_v4.html). It

科学数据来自 [NASA 官方 CSV](https://data.giss.nasa.gov/gistemp/tabledata_v4/GLB.Ts+dSST.csv) 与 [v4 说明](https://data.giss.nasa.gov/gistemp/data_v4.html)，2026-10-23 获取，冻结哈希 `4c64eda288bce32b8d76926278f346ecea1e53506df8ee72a3d5accb0657c416`，各会话离线使用快照

<svg aria-roledescription="flowchart-v2" role="graphics-document document" viewBox="0 0 2918.3515625 3963.89990234375" height="3963.89990234375" class="flowchart" xmlns="http://www.w3.org/2000/svg" width="2918.3515625" id="diagram"><style>#

<svg aria-roledescription="flowchart-v2" role="graphics-document document" viewBox="0 0 2587.296875 3289" height="3289" class="flowchart" xmlns="http://www.w3.org/2000/svg" width="2587.296875" id="diagram"><style>#diagram{font-family:Arial,

<svg aria-roledescription="flowchart-v2" role="graphics-document document" viewBox="0 0 2185.8203125 2731.71875" height="2731.71875" class="flowchart" xmlns="http://www.w3.org/2000/svg" width="2185.8203125" id="diagram"><style>#diagram{font

<svg aria-roledescription="flowchart-v2" role="graphics-document document" viewBox="0 0 2109.79296875 2421.546875" height="2421.546875" class="flowchart" xmlns="http://www.w3.org/2000/svg" width="2109.79296875" id="diagram"><style>#diagram{

<svg aria-roledescription="flowchart-v2" role="graphics-document document" viewBox="0 0 1178.5 1858.328125" height="1858.328125" class="flowchart" xmlns="http://www.w3.org/2000/svg" width="1178.5" id="diagram"><style>#diagram{font-family:Ar

<svg aria-roledescription="flowchart-v2" role="graphics-document document" viewBox="0 0 1150.9649658203125 1626.59375" height="1626.59375" class="flowchart" xmlns="http://www.w3.org/2000/svg" width="1150.9649658203125" id="diagram"><style>#

<svg aria-roledescription="flowchart-v2" role="graphics-document document" viewBox="0 0 1267 7130.5" height="7130.5" class="flowchart" xmlns="http://www.w3.org/2000/svg" width="1267" id="diagram"><style>#diagram{font-family:Arial,"PingFang 

<svg aria-roledescription="flowchart-v2" role="graphics-document document" viewBox="0 0 1217.734375 5702.5" height="5702.5" class="flowchart" xmlns="http://www.w3.org/2000/svg" width="1217.734375" id="diagram"><style>#diagram{font-family:Ar

<svg aria-roledescription="flowchart-v2" role="graphics-document document" viewBox="0 0 1480.25390625 2543.15625" height="2543.15625" class="flowchart" xmlns="http://www.w3.org/2000/svg" width="1480.25390625" id="diagram"><style>#diagram{fo

<svg aria-roledescription="flowchart-v2" role="graphics-document document" viewBox="0 0 1389.2890625 2187.609375" height="2187.609375" class="flowchart" xmlns="http://www.w3.org/2000/svg" width="1389.2890625" id="diagram"><style>#diagram{fo

<svg aria-roledescription="flowchart-v2" role="graphics-document document" viewBox="0 0 896 1925.893798828125" height="1925.893798828125" class="flowchart" xmlns="http://www.w3.org/2000/svg" width="896" id="diagram"><style>#diagram{font-fam

<svg aria-roledescription="flowchart-v2" role="graphics-document document" viewBox="0 0 896 1736.75" height="1736.75" class="flowchart" xmlns="http://www.w3.org/2000/svg" width="896" id="diagram"><style>#diagram{font-family:Arial,"PingFang 

<svg aria-roledescription="sequence" role="graphics-document document" viewBox="-50 -10 1425 1017" height="1017" xmlns="http://www.w3.org/2000/svg" width="1425" id="diagram"><g><rect class="actor actor-bottom" ry="3" rx="3" name="B" height=

<svg aria-roledescription="sequence" role="graphics-document document" viewBox="-50 -10 1067.5 1017" height="1017" xmlns="http://www.w3.org/2000/svg" width="1067.5" id="diagram"><g><rect class="actor actor-bottom" ry="3" rx="3" name="B" hei

<svg aria-roledescription="flowchart-v2" role="graphics-document document" viewBox="0 0 2467.51171875 6767.5546875" height="6767.5546875" class="flowchart" xmlns="http://www.w3.org/2000/svg" width="2467.51171875" id="diagram"><style>#diagra

<svg aria-roledescription="flowchart-v2" role="graphics-document document" viewBox="0 0 2050.390625 5702.97265625" height="5702.97265625" class="flowchart" xmlns="http://www.w3.org/2000/svg" width="2050.390625" id="diagram"><style>#diagram{

<svg aria-roledescription="mindmap" role="graphics-document document" viewBox="-7 -7 2063.592041015625 989.450927734375" height="989.450927734375" xmlns="http://www.w3.org/2000/svg" width="2063.592041015625" id="diagram"><style>#diagram{fon

<svg aria-roledescription="mindmap" role="graphics-document document" viewBox="-7 -7 1848.9188232421875 937.9786376953125" height="937.9786376953125" xmlns="http://www.w3.org/2000/svg" width="1848.9188232421875" id="diagram"><style>#diagram

<svg aria-roledescription="mindmap" role="graphics-document document" viewBox="-7 -7 2241.97314453125 974.725830078125" height="974.725830078125" xmlns="http://www.w3.org/2000/svg" width="2241.97314453125" id="diagram"><style>#diagram{font-

<svg aria-roledescription="mindmap" role="graphics-document document" viewBox="-7 -7 1674.121337890625 959.42529296875" height="959.42529296875" xmlns="http://www.w3.org/2000/svg" width="1674.121337890625" id="diagram"><style>#diagram{font-

<svg aria-roledescription="mindmap" role="graphics-document document" viewBox="-7 -7 2217.490234375 1014.674072265625" height="1014.674072265625" xmlns="http://www.w3.org/2000/svg" width="2217.490234375" id="diagram"><style>#diagram{font-fa

<svg aria-roledescription="mindmap" role="graphics-document document" viewBox="-7 -7 1736.7271728515625 966.61279296875" height="966.61279296875" xmlns="http://www.w3.org/2000/svg" width="1736.7271728515625" id="diagram"><style>#diagram{fon

<svg aria-roledescription="mindmap" role="graphics-document document" viewBox="-7 -7 2528.63525390625 1183.283203125" height="1183.283203125" xmlns="http://www.w3.org/2000/svg" width="2528.63525390625" id="diagram"><style>#diagram{font-fami

<svg aria-roledescription="mindmap" role="graphics-document document" viewBox="-7 -7 1997.1734619140625 1116.6561279296875" height="1116.6561279296875" xmlns="http://www.w3.org/2000/svg" width="1997.1734619140625" id="diagram"><style>#diagr

<svg aria-roledescription="sequence" role="graphics-document document" viewBox="-50 -10 2680.5 6375" height="6375" xmlns="http://www.w3.org/2000/svg" width="2680.5" id="diagram"><g><rect class="actor actor-bottom" ry="3" rx="3" name="B" hei

<svg aria-roledescription="sequence" role="graphics-document document" viewBox="-50 -10 1855.5 6375" height="6375" xmlns="http://www.w3.org/2000/svg" width="1855.5" id="diagram"><g><rect class="actor actor-bottom" ry="3" rx="3" name="B" hei

<svg aria-roledescription="flowchart-v2" role="graphics-document document" viewBox="0 0 1138.75 1703.40625" height="1703.40625" class="flowchart" xmlns="http://www.w3.org/2000/svg" width="1138.75" id="diagram"><style>#diagram{font-family:Ar

<svg aria-roledescription="flowchart-v2" role="graphics-document document" viewBox="0 0 1133.5 1445.21875" height="1445.21875" class="flowchart" xmlns="http://www.w3.org/2000/svg" width="1133.5" id="diagram"><style>#diagram{font-family:Aria

Executable file

| C03 | 应先固定验收，再执行和读回产物 | 能力协议学习单元 1、5、6；[项目研究 §8](../../开源发布/personal-agent-growth/docs/learning-science.zh-CN.md) | 工程验收原则 | 本卡时长任务是设计例子，不是已运行的实验。标准需在结果前固定。 | | C04 | 不能把文件打不开一律判成概念不会 | 能力协议缺口表与失败诊断 | 工程诊断原则 | 环境故障应保留为运行无效或未知；不能借环境标签掩盖已得到

| C04 | 不能把文件打不开一律判成概念不会 | 能力协议缺口表与失败诊断 | 工程诊断原则 | 环境故障应保留为运行无效或未知；不能借环境标签掩盖已得到的错误产物。 | | C05 | 重复提取在人类词汇实验中改善延迟回忆 | [Karpicke & Roediger，2008，原始研究摘要](https://pubmed.ncbi.nlm.nih.gov/18276894/) | 有限定的研究事实 | 对象为学生、任务为外语词汇；重复阅读在该设计中未改善延迟回忆，不推

| C05 | 重复提取在人类词汇实验中改善延迟回忆 | [Karpicke & Roediger，2008，原始研究摘要](https://pubmed.ncbi.nlm.nih.gov/18276894/) | 有限定的研究事实 | 对象为学生、任务为外语词汇；重复阅读在该设计中未改善延迟回忆，不推广到所有学习、任务或 Agent。DOI 直连 403 已保留。 |

| C06 | 控制间隔的后续实验中，重读也改善学习 | [Soderstrom 等，2016，原始研究摘要](https://pubmed.ncbi.nlm.nih.gov/26674128/) | 有限定的研究事实 | 摘要报告实验 1 复现原结果，实验 2 在被试内操纵并控制间隔后两种方式均有益；不能因此宣布二者始终等效。改变了我对“提取一般优于重读”的初判。 |

| C07 | 自我解释不保证优于重读 | [Harders & Ebersbach，2026，摘要](https://onlinelibrary.wiley.com/doi/10.1002/acp.70174) | 有限定的研究事实 | 预注册 N=208、虚构事实、立即与两周后测试未检出策略优势；不是所有任务无效的证明，也不能把未检出差异解释成严格等效。 |

| C08 | 自我解释与表现有关不自动证明因果 | [Chi 等，1989，原始研究摘要](https://onlinelibrary.wiley.com/doi/abs/10.1207/s15516709cog1302_1) | 来源观察事实及证据范围判断 | 摘要比较好/差学习者的力学例题解释与解题，不足以认定已随机控制所有差异；本轮未读其全部方法，不声称完成设计审计。其 AI 理论讨论不是对现代冻结模型的验证。 |

人类实验提供线索，不能替 Agent 验收。[词汇研究](https://pubmed.ncbi.nlm.nih.gov/18276894/)发现重复提取有利于延迟回忆；[后续研究](https://pubmed.ncbi.nlm.nih.gov/26674128/)控制间隔后，重读也有收益。[另一项实验](https://onlinelibrary.wiley.com/doi/10.1002/acp.70174)在虚构事实任务中未发现自我解释优于重读。不能把这些结果写成“多

|---|---| | [项目教育研究](../../开源发布/personal-agent-growth/docs/learning-science.zh-CN.md)、[能力协议](../../开源发布/personal-agent-growth/references/capability-learning.md) | 已全文读取；固定权重范围、工程假设与分维验收的项目规则 | | [Karpicke & Roediger，2008](https://pubmed.ncb

| [项目教育研究](../../开源发布/personal-agent-growth/docs/learning-science.zh-CN.md)、[能力协议](../../开源发布/personal-agent-growth/references/capability-learning.md) | 已全文读取；固定权重范围、工程假设与分维验收的项目规则 | | [Karpicke & Roediger，2008](https://pubmed.ncbi.nlm.nih.

| [Karpicke & Roediger，2008](https://pubmed.ncbi.nlm.nih.gov/18276894/) | 原始研究摘要；DOI 直接访问 403 后，读取 PubMed 收录的作者摘要 |

| [Soderstrom、Kerr & Bjork，2016](https://pubmed.ncbi.nlm.nih.gov/26674128/) | 原始研究摘要；间隔条件改变了重读的结果 |

| [Harders & Ebersbach，2026](https://onlinelibrary.wiley.com/doi/10.1002/acp.70174) | 原始论文摘要与背景；未完成方法、数据与统计全量审查 |

| [Chi 等，1989](https://onlinelibrary.wiley.com/doi/abs/10.1207/s15516709cog1302_1) | 原始研究摘要；用于审查自我解释主张的边界，正文不展开 |

[Credential-related source omitted]

[Credential-related source omitted]

[Credential-related source omitted]

[Credential-related source omitted]

Executable file

[Credential-related source omitted]

[Credential-related source omitted]

... # Subtest: CMD and shell deadlines do not pass overflowing timers to Windows curl ok 1414 - CMD and shell deadlines do not pass overflowing timers to Windows curl

# Subtest: CMD and shell deadlines do not pass overflowing timers to Windows curl ok 1414 - CMD and shell deadlines do not pass overflowing timers to Windows curl   ---

ok 1414 - CMD and shell deadlines do not pass overflowing timers to Windows curl   ---   duration_ms: 0.272833

[Credential-related source omitted]

[Credential-related source omitted]

... # Subtest: CMD and shell deadlines do not pass overflowing timers to Windows curl ok 1414 - CMD and shell deadlines do not pass overflowing timers to Windows curl

# Subtest: CMD and shell deadlines do not pass overflowing timers to Windows curl ok 1414 - CMD and shell deadlines do not pass overflowing timers to Windows curl   ---

ok 1414 - CMD and shell deadlines do not pass overflowing timers to Windows curl   ---   duration_ms: 0.613792

Executable file

Unsupported or unresolved allowed-tools declaration

"conclusion": "success",   "headSha": "52caa11c947f5bc9e5422c8798801ce0ed6d8083",   "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40387260224",

"headSha": "52caa11c947f5bc9e5422c8798801ce0ed6d8083",   "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40387260224",   "jobs": [

"url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40387260224",

"startedAt": "2026-10-26T19:29:58Z",       "completedAt": "2026-10-26T19:30:33Z",       "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40387260224/job/121798727318"

"completedAt": "2026-10-26T19:30:33Z",       "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40387260224/job/121798727318"     },

"url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40387260224/job/121798727318"

"startedAt": "2026-10-26T19:29:55Z",       "completedAt": "2026-10-26T19:30:30Z",       "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40387260224/job/121798727593"

"completedAt": "2026-10-26T19:30:30Z",       "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40387260224/job/121798727593"     },

"url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40387260224/job/121798727593"

"startedAt": "2026-10-26T19:29:55Z",       "completedAt": "2026-10-26T19:30:03Z",       "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40387260224/job/121798727672"

"completedAt": "2026-10-26T19:30:03Z",       "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40387260224/job/121798727672"     }

"url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40387260224/job/121798727672"

"full_log_sha256": "8d747d026d44238bce48c228b3f7f6898930e89806f93f5ff1670ba104c2e884",   "prior_failure": {     "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40387006023",

"prior_failure": {     "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40387006023",     "commit": "826d0c335e22201575ffc68b85586b9e27696b2d",

"url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40387006023",

{   "version": "0.4.1",   "workflow_url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40387005475",

"version": "0.4.1",   "workflow_url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40387005475",   "results": [

"workflow_url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40387005475",

}         ],         "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40395595837/job/121827649916"

],         "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40395595837/job/121827649916"       },

"url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40395595837/job/121827649916"

}         ],         "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40395595837/job/121827650141"

],         "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40395595837/job/121827650141"       },

"url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40395595837/job/121827650141"

}         ],         "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40395595837/job/121827650432"

],         "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40395595837/job/121827650432"       }

"url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40395595837/job/121827650432"

],     "status": "completed",     "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40395595837"

"status": "completed",     "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40395595837"   },

"url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40395595837"

}         ],         "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40395525859/job/121827404643"

],         "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40395525859/job/121827404643"       },

"url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40395525859/job/121827404643"

}         ],         "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40395525859/job/121827404960"

],         "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40395525859/job/121827404960"       }

"url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40395525859/job/121827404960"

],     "status": "completed",     "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40395525859"

"status": "completed",     "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40395525859"   },

"url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40395525859"

"prior_failures": [     {       "run": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40394859606",

{       "run": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40394859606",       "cause": "The Windows guide transport adapter used hard-coded 0.4.1 filenames and download directory.",

"run": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40394859606",       "cause": "The Windows guide transport adapter used hard-coded 0.4.1 filenames and download directory.",       "repair": "Derive current version fro

},     {       "run": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40395526150",

{       "run": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40395526150",       "cause": "New regression compared an 8.3 Windows temp path against its resolved long spelling.",

"run": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40395526150",

}           ],           "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40499055634/job/122171827523"

],           "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40499055634/job/122171827523"         },

"url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40499055634/job/122171827523"

}           ],           "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40499055634/job/122171827807"

],           "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40499055634/job/122171827807"         },

"url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40499055634/job/122171827807"

}           ],           "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40499055634/job/122171827946"

],           "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40499055634/job/122171827946"         }

"url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40499055634/job/122171827946"

],       "status": "completed",       "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40499055634",

"status": "completed",       "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40499055634",       "raw_log_sha256": "fb3ec809f4721b02aefa74dfe7fd7fdfea29838c2b4750d02f8be918d207d787"

"url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40499055634",

}           ],           "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40499104681/job/122172001207"

],           "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40499104681/job/122172001207"         },

"url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40499104681/job/122172001207"

}           ],           "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40499104681/job/122172001593"

],           "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40499104681/job/122172001593"         }

"url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40499104681/job/122172001593"

],       "status": "completed",       "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40499104681",

"status": "completed",       "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40499104681",       "raw_log_sha256": "e49ddadb1cadda979723f631ecca14ce3e1ef41ea236fd5d07da995f16194251"

"url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40499104681",

{         "file": "personal-agent-growth/references/regression-repair.en.md",         "source": "https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.3/docs/evidence/koder-skip-audit-20261027.json",

"file": "personal-agent-growth/references/regression-repair.en.md",         "source": "https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.3/docs/evidence/koder-skip-audit-20261027.json",         "source_sha256": "d3bb78bb068d6a82

"source": "https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.3/docs/evidence/koder-skip-audit-20261027.json",

{         "file": "personal-agent-growth/references/regression-repair.en.md",         "source": "https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.2/docs/evidence/koder-regression-repair-20261027.json",

"file": "personal-agent-growth/references/regression-repair.en.md",         "source": "https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.2/docs/evidence/koder-regression-repair-20261027.json",         "source_sha256": "13a50de97

"source": "https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.2/docs/evidence/koder-regression-repair-20261027.json",

{         "file": "personal-agent-growth/references/regression-repair.en.md",         "source": "https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.2/docs/release-0.4.2-acceptance.en.md",

"file": "personal-agent-growth/references/regression-repair.en.md",         "source": "https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.2/docs/release-0.4.2-acceptance.en.md",         "source_sha256": "4535e9447c59fae617b0d2ac9

"source": "https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.2/docs/release-0.4.2-acceptance.en.md",

{         "file": "personal-agent-growth/references/regression-repair.md",         "source": "https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.3/docs/evidence/koder-skip-audit-20261027.json",

"file": "personal-agent-growth/references/regression-repair.md",         "source": "https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.3/docs/evidence/koder-skip-audit-20261027.json",         "source_sha256": "d3bb78bb068d6a827a4

"source": "https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.3/docs/evidence/koder-skip-audit-20261027.json",

{         "file": "personal-agent-growth/references/regression-repair.md",         "source": "https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.2/docs/evidence/koder-regression-repair-20261027.json",

"file": "personal-agent-growth/references/regression-repair.md",         "source": "https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.2/docs/evidence/koder-regression-repair-20261027.json",         "source_sha256": "13a50de9751c

"source": "https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.2/docs/evidence/koder-regression-repair-20261027.json",

{         "file": "personal-agent-growth/references/regression-repair.md",         "source": "https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.2/docs/release-0.4.2-acceptance.zh-CN.md",

"file": "personal-agent-growth/references/regression-repair.md",         "source": "https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.2/docs/release-0.4.2-acceptance.zh-CN.md",         "source_sha256": "7bf57feb7f56889de84ef4f52

"source": "https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.2/docs/release-0.4.2-acceptance.zh-CN.md",

"status": "PASS",   "source_commit": "8b662fae7a47534295ae1a4ed6d51f578f82a7f3",   "workflow": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40499104681",

"source_commit": "8b662fae7a47534295ae1a4ed6d51f578f82a7f3",   "workflow": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40499104681",   "scope": "Public installer and shared-file processes on hosted Windows; separate fr

"workflow": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40499104681",

}           ],           "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40645260382/job/122672483499"

],           "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40645260382/job/122672483499"         },

"url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40645260382/job/122672483499"

}           ],           "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40645260382/job/122672483737"

],           "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40645260382/job/122672483737"         },

"url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40645260382/job/122672483737"

}           ],           "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40645260382/job/122672483770"

],           "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40645260382/job/122672483770"         }

"url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40645260382/job/122672483770"

],       "status": "completed",       "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40645260382",

"status": "completed",       "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40645260382",       "raw_log_sha256": "9ffee7a5b6984b4d910d58632b3121a5c7b20906ba0ab06d6e146fbf28b55663"

"url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40645260382",

}           ],           "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40645260603/job/122672484071"

],           "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40645260603/job/122672484071"         },

"url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40645260603/job/122672484071"

}           ],           "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40645260603/job/122672484304"

],           "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40645260603/job/122672484304"         }

"url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40645260603/job/122672484304"

],       "status": "completed",       "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40645260603",

"status": "completed",       "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40645260603",       "raw_log_sha256": "9891007f579156be198d79a2720535f63c34006a9a68d0bd14a46b20320369f0"

"url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40645260603",

{         "file": "personal-agent-growth/references/regression-repair.en.md",         "source": "https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.3/docs/evidence/koder-skip-audit-20261027.json",

"file": "personal-agent-growth/references/regression-repair.en.md",         "source": "https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.3/docs/evidence/koder-skip-audit-20261027.json",         "source_sha256": "d3bb78bb068d6a82

"source": "https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.3/docs/evidence/koder-skip-audit-20261027.json",

{         "file": "personal-agent-growth/references/regression-repair.en.md",         "source": "https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.2/docs/evidence/koder-regression-repair-20261027.json",

"file": "personal-agent-growth/references/regression-repair.en.md",         "source": "https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.2/docs/evidence/koder-regression-repair-20261027.json",         "source_sha256": "13a50de97

"source": "https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.2/docs/evidence/koder-regression-repair-20261027.json",

{         "file": "personal-agent-growth/references/regression-repair.en.md",         "source": "https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.2/docs/release-0.4.2-acceptance.en.md",

"file": "personal-agent-growth/references/regression-repair.en.md",         "source": "https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.2/docs/release-0.4.2-acceptance.en.md",         "source_sha256": "4535e9447c59fae617b0d2ac9

"source": "https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.2/docs/release-0.4.2-acceptance.en.md",

{         "file": "personal-agent-growth/references/regression-repair.md",         "source": "https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.3/docs/evidence/koder-skip-audit-20261027.json",

"file": "personal-agent-growth/references/regression-repair.md",         "source": "https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.3/docs/evidence/koder-skip-audit-20261027.json",         "source_sha256": "d3bb78bb068d6a827a4

"source": "https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.3/docs/evidence/koder-skip-audit-20261027.json",

{         "file": "personal-agent-growth/references/regression-repair.md",         "source": "https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.2/docs/evidence/koder-regression-repair-20261027.json",

"file": "personal-agent-growth/references/regression-repair.md",         "source": "https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.2/docs/evidence/koder-regression-repair-20261027.json",         "source_sha256": "13a50de9751c

"source": "https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.2/docs/evidence/koder-regression-repair-20261027.json",

{         "file": "personal-agent-growth/references/regression-repair.md",         "source": "https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.2/docs/release-0.4.2-acceptance.zh-CN.md",

"file": "personal-agent-growth/references/regression-repair.md",         "source": "https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.2/docs/release-0.4.2-acceptance.zh-CN.md",         "source_sha256": "7bf57feb7f56889de84ef4f52

"source": "https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.2/docs/release-0.4.2-acceptance.zh-CN.md",

"status": "PASS",   "source_commit": "81876f18363cfe3de767ba5e8028fb60a819a43b",   "workflow": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40645260603",

"source_commit": "81876f18363cfe3de767ba5e8028fb60a819a43b",   "workflow": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40645260603",   "scope": "Hosted Windows install, shared-file and removal checks in PowerShell 5.1 

"workflow": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40645260603",

TAP version 13 # \#< CLIXML # <Objs Version="1.1.0.1" xmlns="http://schemas.microsoft.com/powershell/2004/04"><Obj S="progress" RefId="0"><TN RefId="0"><T>System.Management.Automation.PSCustomObject</T><T>System.Object</T></TN><MS><I64 N="S

# \#< CLIXML # <Objs Version="1.1.0.1" xmlns="http://schemas.microsoft.com/powershell/2004/04"><Obj S="progress" RefId="0"><TN RefId="0"><T>System.Management.Automation.PSCustomObject</T><T>System.Object</T></TN><MS><I64 N="SourceId">1</I64

# <Objs Version="1.1.0.1" xmlns="http://schemas.microsoft.com/powershell/2004/04"><Obj S="progress" RefId="0"><TN RefId="0"><T>System.Management.Automation.PSCustomObject</T><T>System.Object</T></TN><MS><I64 N="SourceId">1</I64><PR N="Recor

# <Objs Version="1.1.0.1" xmlns="http://schemas.microsoft.com/powershell/2004/04"><Obj S="progress" RefId="0"><TN RefId="0"><T>System.Management.Automation.PSCustomObject</T><T>System.Object</T></TN><MS><I64 N="SourceId">1</I64><PR N="Recor

# <Objs Version="1.1.0.1" xmlns="http://schemas.microsoft.com/powershell/2004/04"><Obj S="progress" RefId="0"><TN RefId="0"><T>System.Management.Automation.PSCustomObject</T><T>System.Object</T></TN><MS><I64 N="SourceId">1</I64><PR N="Recor

type: 'test'   ... # <Objs Version="1.1.0.1" xmlns="http://schemas.microsoft.com/powershell/2004/04"><Obj S="progress" RefId="0"><TN RefId="0"><T>System.Management.Automation.PSCustomObject</T><T>System.Object</T></TN><MS><I64 N="SourceId">

... # <Objs Version="1.1.0.1" xmlns="http://schemas.microsoft.com/powershell/2004/04"><Obj S="progress" RefId="0"><TN RefId="0"><T>System.Management.Automation.PSCustomObject</T><T>System.Object</T></TN><MS><I64 N="SourceId">1</I64><PR N="R

# <Objs Version="1.1.0.1" xmlns="http://schemas.microsoft.com/powershell/2004/04"><Obj S="progress" RefId="0"><TN RefId="0"><T>System.Management.Automation.PSCustomObject</T><T>System.Object</T></TN><MS><I64 N="SourceId">1</I64><PR N="Recor

TAP version 13 # \#< CLIXML # <Objs Version="1.1.0.1" xmlns="http://schemas.microsoft.com/powershell/2004/04"><Obj S="progress" RefId="0"><TN RefId="0"><T>System.Management.Automation.PSCustomObject</T><T>System.Object</T></TN><MS><I64 N="S

# \#< CLIXML # <Objs Version="1.1.0.1" xmlns="http://schemas.microsoft.com/powershell/2004/04"><Obj S="progress" RefId="0"><TN RefId="0"><T>System.Management.Automation.PSCustomObject</T><T>System.Object</T></TN><MS><I64 N="SourceId">1</I64

# <Objs Version="1.1.0.1" xmlns="http://schemas.microsoft.com/powershell/2004/04"><Obj S="progress" RefId="0"><TN RefId="0"><T>System.Management.Automation.PSCustomObject</T><T>System.Object</T></TN><MS><I64 N="SourceId">1</I64><PR N="Recor

# <Objs Version="1.1.0.1" xmlns="http://schemas.microsoft.com/powershell/2004/04"><Obj S="progress" RefId="0"><TN RefId="0"><T>System.Management.Automation.PSCustomObject</T><T>System.Object</T></TN><MS><I64 N="SourceId">1</I64><PR N="Recor

# <Objs Version="1.1.0.1" xmlns="http://schemas.microsoft.com/powershell/2004/04"><Obj S="progress" RefId="0"><TN RefId="0"><T>System.Management.Automation.PSCustomObject</T><T>System.Object</T></TN><MS><I64 N="SourceId">1</I64><PR N="Recor

type: 'test'   ... # <Objs Version="1.1.0.1" xmlns="http://schemas.microsoft.com/powershell/2004/04"><Obj S="progress" RefId="0"><TN RefId="0"><T>System.Management.Automation.PSCustomObject</T><T>System.Object</T></TN><MS><I64 N="SourceId">

... # <Objs Version="1.1.0.1" xmlns="http://schemas.microsoft.com/powershell/2004/04"><Obj S="progress" RefId="0"><TN RefId="0"><T>System.Management.Automation.PSCustomObject</T><T>System.Object</T></TN><MS><I64 N="SourceId">1</I64><PR N="R

# <Objs Version="1.1.0.1" xmlns="http://schemas.microsoft.com/powershell/2004/04"><Obj S="progress" RefId="0"><TN RefId="0"><T>System.Management.Automation.PSCustomObject</T><T>System.Object</T></TN><MS><I64 N="SourceId">1</I64><PR N="Recor

#!/usr/bin/env python3

#!/usr/bin/env python3

Project sources, inspected for this workshop: [adaptive tutorial](https://github.com/KarmenLoop/personal-agent-growth/blob/main/docs/adaptive-learning.en.md), [adaptive protocol](https://github.com/KarmenLoop/personal-agent-growth/blob/main

为本工作坊检查的项目来源：[自适应教程](https://github.com/KarmenLoop/personal-agent-growth/blob/main/docs/adaptive-learning.en.md)、[自适应协议](https://github.com/KarmenLoop/personal-agent-growth/blob/main/references/adaptive-learning.en.md)、[能力协议](https://github

Unsupported or unresolved allowed-tools declaration

Unsupported or unresolved allowed-tools declaration

Unsupported or unresolved allowed-tools declaration

Unsupported or unresolved allowed-tools declaration

Unsupported or unresolved allowed-tools declaration

Unsupported or unresolved allowed-tools declaration

Unsupported or unresolved allowed-tools declaration

Unsupported or unresolved allowed-tools declaration

Unsupported or unresolved allowed-tools declaration

Unsupported or unresolved allowed-tools declaration

Unsupported or unresolved allowed-tools declaration

Unsupported or unresolved allowed-tools declaration

Unsupported or unresolved allowed-tools declaration

<?xml version="1.0" encoding="utf-8" standalone="no"?> <!DOCTYPE svg PUBLIC "-//W3C//DTD SVG 1.1//EN"   "http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd">

<!DOCTYPE svg PUBLIC "-//W3C//DTD SVG 1.1//EN"   "http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd"> <svg xmlns:xlink="http://www.w3.org/1999/xlink" width="864pt" height="576pt" viewBox="0 0 864 576" xmlns="http://www.w3.org/2000/svg" versi

"http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd">

<svg xmlns:xlink="http://www.w3.org/1999/xlink" width="864pt" height="576pt" viewBox="0 0 864 576" xmlns="http://www.w3.org/2000/svg" version="1.1">

<metadata>   <rdf:RDF xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:cc="http://creativecommons.org/ns#" xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#">    <cc:Work>

<rdf:RDF xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:cc="http://creativecommons.org/ns#" xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#">

<cc:Work>     <dc:type rdf:resource="http://purl.org/dc/dcmitype/StillImage"/>     <dc:date>2026-10-23T16:24:03.664643</dc:date>

<dc:type rdf:resource="http://purl.org/dc/dcmitype/StillImage"/>

<dc:creator>      <cc:Agent>       <dc:title>Matplotlib v3.10.8, https://matplotlib.org/</dc:title>

<cc:Agent>       <dc:title>Matplotlib v3.10.8, https://matplotlib.org/</dc:title>      </cc:Agent>

<dc:title>Matplotlib v3.10.8, https://matplotlib.org/</dc:title>

<svg xmlns="http://www.w3.org/2000/svg" width="760" height="395" viewBox="0 0 760 395"><title>Personal Agent Growth v0.3.3</title><desc>Explanatory workflow diagram for version 0.3.3; not a host screenshot or runtime result.</desc><rect wid

<svg xmlns="http://www.w3.org/2000/svg" width="760" height="395" viewBox="0 0 760 395"><title>Personal Agent Growth v0.4.0</title><desc>Explanatory workflow diagram for version 0.4.0; not a host screenshot or runtime result.</desc><rect wid

SkillEntry="$GrowthSkillsRoot/personal-agent-growth"   ReleaseDir="$GrowthRoot/downloads/0.7.0"   ReleaseUrl="https://github.com/KarmenLoop/personal-agent-growth/releases/download/v0.7.0"

ReleaseDir="$GrowthRoot/downloads/0.7.0"   ReleaseUrl="https://github.com/KarmenLoop/personal-agent-growth/releases/download/v0.7.0"   if [ -e "$SkillDir" ] || [ -L "$SkillDir" ] || [ -e "$SkillEntry" ] || [ -L "$SkillEntry" ]; then

ReleaseUrl="https://github.com/KarmenLoop/personal-agent-growth/releases/download/v0.7.0"

fi   mkdir -p "$ReleaseDir" "$GrowthRoot/skills" "$GrowthRoot/tasks" "$GrowthSkillsRoot"   curl --fail --location --retry 2 --max-time 90 "$ReleaseUrl/personal-agent-growth-0.7.0.zip" -o "$ReleaseDir/personal-agent-growth-0.7.0.zip"

mkdir -p "$ReleaseDir" "$GrowthRoot/skills" "$GrowthRoot/tasks" "$GrowthSkillsRoot"   curl --fail --location --retry 2 --max-time 90 "$ReleaseUrl/personal-agent-growth-0.7.0.zip" -o "$ReleaseDir/personal-agent-growth-0.7.0.zip"   curl --fai

curl --fail --location --retry 2 --max-time 90 "$ReleaseUrl/personal-agent-growth-0.7.0.zip" -o "$ReleaseDir/personal-agent-growth-0.7.0.zip"

curl --fail --location --retry 2 --max-time 90 "$ReleaseUrl/manifest.json" -o "$ReleaseDir/manifest.json"

$SkillEntry = Join-Path $GrowthSkillsRoot 'personal-agent-growth'   $ReleaseDir = Join-Path $GrowthRoot 'downloads\0.7.0'   $ReleaseUrl = 'https://github.com/KarmenLoop/personal-agent-growth/releases/download/v0.7.0'

$ReleaseDir = Join-Path $GrowthRoot 'downloads\0.7.0'   $ReleaseUrl = 'https://github.com/KarmenLoop/personal-agent-growth/releases/download/v0.7.0'   if ((Test-Path -LiteralPath $SkillDir) -or (Get-Item -LiteralPath $SkillEntry -Force -Err

$ReleaseUrl = 'https://github.com/KarmenLoop/personal-agent-growth/releases/download/v0.7.0'

Download the exact asset and its manifest from [v0.7.0](https://github.com/KarmenLoop/personal-agent-growth/releases/tag/v0.7.0). GitHub's automatic Source code ZIP serves a different purpose from the installer asset. Extract the whole fold

Directory conventions are checked against the [Agent Skills specification](https://agentskills.io/specification), [Cursor Skills documentation](https://cursor.com/docs/skills), and actual host capabilities. Preserve [LICENSE](../LICENSE) an

SkillEntry="$GrowthSkillsRoot/personal-agent-growth"   ReleaseDir="$GrowthRoot/downloads/0.7.0"   ReleaseUrl="https://github.com/KarmenLoop/personal-agent-growth/releases/download/v0.7.0"

ReleaseDir="$GrowthRoot/downloads/0.7.0"   ReleaseUrl="https://github.com/KarmenLoop/personal-agent-growth/releases/download/v0.7.0"   if [ -e "$SkillDir" ] || [ -L "$SkillDir" ] || [ -e "$SkillEntry" ] || [ -L "$SkillEntry" ]; then

ReleaseUrl="https://github.com/KarmenLoop/personal-agent-growth/releases/download/v0.7.0"

fi   mkdir -p "$ReleaseDir" "$GrowthRoot/skills" "$GrowthRoot/tasks" "$GrowthSkillsRoot"   curl --fail --location --retry 2 --max-time 90 "$ReleaseUrl/personal-agent-growth-0.7.0.zip" -o "$ReleaseDir/personal-agent-growth-0.7.0.zip"

mkdir -p "$ReleaseDir" "$GrowthRoot/skills" "$GrowthRoot/tasks" "$GrowthSkillsRoot"   curl --fail --location --retry 2 --max-time 90 "$ReleaseUrl/personal-agent-growth-0.7.0.zip" -o "$ReleaseDir/personal-agent-growth-0.7.0.zip"   curl --fai

curl --fail --location --retry 2 --max-time 90 "$ReleaseUrl/personal-agent-growth-0.7.0.zip" -o "$ReleaseDir/personal-agent-growth-0.7.0.zip"

curl --fail --location --retry 2 --max-time 90 "$ReleaseUrl/manifest.json" -o "$ReleaseDir/manifest.json"

$SkillEntry = Join-Path $GrowthSkillsRoot 'personal-agent-growth'   $ReleaseDir = Join-Path $GrowthRoot 'downloads\0.7.0'   $ReleaseUrl = 'https://github.com/KarmenLoop/personal-agent-growth/releases/download/v0.7.0'

$ReleaseDir = Join-Path $GrowthRoot 'downloads\0.7.0'   $ReleaseUrl = 'https://github.com/KarmenLoop/personal-agent-growth/releases/download/v0.7.0'   if ((Test-Path -LiteralPath $SkillDir) -or (Get-Item -LiteralPath $SkillEntry -Force -Err

$ReleaseUrl = 'https://github.com/KarmenLoop/personal-agent-growth/releases/download/v0.7.0'

从 [v0.7.0](https://github.com/KarmenLoop/personal-agent-growth/releases/tag/v0.7.0)下载名称准确的附件与对应 manifest。GitHub 自动生成的 Source code ZIP 与安装附件用途不同。解压整个文件夹到 `.agents/skills/personal-agent-growth`；应直接看到 `SKILL.md`，不要多套一层目录。

安装目录依据 [Agent Skills 规范](https://agentskills.io/specification)、[Cursor Skills 文档](https://cursor.com/docs/skills)与宿主实际能力核对。保留 [LICENSE](../LICENSE) 与 [NOTICE](../NOTICE)。发布验收与历史证据见[兼容性](compatibility.zh-CN.md)。

At university level, [MIT's backward-design guidance](https://tll.mit.edu/teaching-resources/course-design/backward-design/) begins with intended outcomes and acceptable evidence before selecting activities. [CMU's learning principles](http

Richard Feynman was a physicist and teacher. His [preface to the Lectures](https://www.feynmanlectures.caltech.edu/I_91.html) describes lectures as part of a course that also included discussion and laboratory work, and discusses limitation

[Caltech's current teaching-to-learn guidance](https://ctlo.caltech.edu/aboutctlo/whoweserve/undergraduates/learning-resources/learning/power-of-teaching) recommends explaining without notes, locating gaps, returning to the material, and en

Explanation is a diagnostic tool, not automatic proof of learning. In a [2026 preregistered study](https://onlinelibrary.wiley.com/doi/10.1002/acp.70174), 208 participants learned fictional facts. The reported comparison did not find a lear

| ID | Source | Material reviewed and access | |---|---|---| | E01 | [IES / WWC, Organizing Instruction and Study to Improve Student Learning (2007)](https://ies.ed.gov/ncee/wwc/PracticeGuide/1) | Official recommendations page opened; evide

|---|---|---| | E01 | [IES / WWC, Organizing Instruction and Study to Improve Student Learning (2007)](https://ies.ed.gov/ncee/wwc/PracticeGuide/1) | Official recommendations page opened; evidence ratings differ across recommendations | | E

| E01 | [IES / WWC, Organizing Instruction and Study to Improve Student Learning (2007)](https://ies.ed.gov/ncee/wwc/PracticeGuide/1) | Official recommendations page opened; evidence ratings differ across recommendations |

| E02 | [Karpicke & Roediger, The Critical Importance of Retrieval for Learning, Science (2008)](https://doi.org/10.1126/science.1152408) | Original article abstract; human vocabulary-learning experiment |

| E03 | [Chi et al., Self-Explanations, Cognitive Science (1989)](https://onlinelibrary.wiley.com/doi/abs/10.1207/s15516709cog1302_1) | Original article abstract; mechanics examples and subsequent problem solving |

| E04 | [Collins, Brown & Holum, Cognitive Apprenticeship (1991)](https://www.aft.org/ae/winter1991/collins_brown_holum) | Authors' article, including methods and their limits |

| E05 | [Feynman, Preface to The Feynman Lectures on Physics (1963)](https://www.feynmanlectures.caltech.edu/I_91.html) | Volume I page body; a Volume III attempt first returned a JavaScript notice |

| E06 | [Caltech CTLO, The Power of Teaching](https://ctlo.caltech.edu/aboutctlo/whoweserve/undergraduates/learning-resources/learning/power-of-teaching) | Official guidance and its attribution |

| E07 | [Singapore MOE, Primary Mathematics Syllabus](https://www.moe.gov.sg/-/media/files/primary/mathematics_syllabus_primary_1_to_6-) and [Secondary Mathematics Syllabus (2020)](https://www.moe.gov.sg/-/media/files/secondary/syllabuses/m

| E08 | [Finnish National Agency for Education, Core Curriculum for Basic Education](https://www.oph.fi/en/education-and-qualifications/national-core-curriculum-primary-and-lower-secondary-basic-education) | Official page body |

| E09 | [Estonia Ministry of Education and Research, General Education](https://hm.ee/en/education-research-and-youth-affairs/general-education/general-education-estonia) | Official page describing the framework introduced in 2014; not a cu

| E10 | [Japan NIER, Science Education at the Stage of Compulsory Education in Japan (2021)](https://www.nier.go.jp/English/educationjapan/pdf/20210623-02.pdf) | Official PDF, especially pages 16–18 |

| E11 | [MIT Teaching + Learning Lab, Backward Design](https://tll.mit.edu/teaching-resources/course-design/backward-design/) | Official indexed text; direct retrieval returned HTTP 502 |

| E12 | [CMU Eberly Center, Learning Principles](https://www.cmu.edu/teaching/principles/learning.html) | Official page body |

| E13 | [Harvard Project Zero, Teaching for Understanding](https://pz.gse.harvard.edu/projects/teaching-for-understanding) | Official indexed text; direct retrieval returned HTTP 502 |

| E14 | [Harvard Project Zero, Visible Thinking](https://pz.gse.harvard.edu/projects/visible-thinking) | Official indexed text |

| E15 | [Harders & Ebersbach, No Causal Self-Explanation Effect for Factual Knowledge (2026)](https://onlinelibrary.wiley.com/doi/10.1002/acp.70174) | Original article abstract and relevant background; no full methods or analysis review cla

大学层面，[MIT 的反向设计](https://tll.mit.edu/teaching-resources/course-design/backward-design/)先明确学习结果和可接受证据，再设计活动；[CMU 的学习原则](https://www.cmu.edu/teaching/principles/learning.html)区分组件技能、技能整合以及何时适用；[Harvard Project Zero 的理解教学](https://pz.gse.harva

Richard Feynman 是物理学家和教师。[《费曼物理学讲义》序言](https://www.feynmanlectures.caltech.edu/I_91.html)介绍的课程不仅包含讲授，也有讨论和实验；作者同时讨论教学尝试的不足。这不是“费曼心理学”，也不构成通用四步学习定律。

[Caltech 现行教学资源](https://ctlo.caltech.edu/aboutctlo/whoweserve/undergraduates/learning-resources/learning/power-of-teaching)建议放下笔记解释、定位漏洞、返回材料补缺，并尽可能面向真实受众。该页面明确说明其内容改编自其他学习资源。我们保留这一出处，不把所有步骤都说成费曼首创。

解释适合诊断，但不能自动证明学习。[2026 年一项预注册研究](https://onlinelibrary.wiley.com/doi/10.1002/acp.70174)让 208 名参与者学习虚构事实，报告未发现自我解释相对重读在立即测试或两周后测试中的策略优势；正确推论与表现相关。这个结果只覆盖该研究任务，不能扩大成“自我解释一律无效”。本轮读取了摘要和相关背景，没有声称完成全部方法与统计分析审查。

| ID | 来源 | 本次读取范围与状态 | |---|---|---| | E01 | [IES / WWC, Organizing Instruction and Study to Improve Student Learning (2007)](https://ies.ed.gov/ncee/wwc/PracticeGuide/1) | 官方建议页已打开；不同建议的证据等级有别 |

|---|---|---| | E01 | [IES / WWC, Organizing Instruction and Study to Improve Student Learning (2007)](https://ies.ed.gov/ncee/wwc/PracticeGuide/1) | 官方建议页已打开；不同建议的证据等级有别 | | E02 | [Karpicke & Roediger, The Critical Importance of Retrieval 

| E01 | [IES / WWC, Organizing Instruction and Study to Improve Student Learning (2007)](https://ies.ed.gov/ncee/wwc/PracticeGuide/1) | 官方建议页已打开；不同建议的证据等级有别 |

| E02 | [Karpicke & Roediger, The Critical Importance of Retrieval for Learning, Science (2008)](https://doi.org/10.1126/science.1152408) | 原始论文摘要；人类词汇学习实验 |

| E03 | [Chi et al., Self-Explanations, Cognitive Science (1989)](https://onlinelibrary.wiley.com/doi/abs/10.1207/s15516709cog1302_1) | 原始论文摘要；力学例题学习与后续解题 |

| E04 | [Collins, Brown & Holum, Cognitive Apprenticeship (1991)](https://www.aft.org/ae/winter1991/collins_brown_holum) | 作者文章，包括方法和限制的正文段落 |

| E05 | [Feynman, Preface to The Feynman Lectures on Physics (1963)](https://www.feynmanlectures.caltech.edu/I_91.html) | I 卷页面正文；此前 III 卷直接访问曾只返回 JavaScript 提示 |

| E06 | [Caltech CTLO, The Power of Teaching](https://ctlo.caltech.edu/aboutctlo/whoweserve/undergraduates/learning-resources/learning/power-of-teaching) | 官方正文和改编出处 |

| E07 | [Singapore MOE, Primary Mathematics Syllabus](https://www.moe.gov.sg/-/media/files/primary/mathematics_syllabus_primary_1_to_6-)；[Secondary Mathematics Syllabus (2020)](https://www.moe.gov.sg/-/media/files/secondary/syllabuses/maths

| E08 | [Finnish National Agency for Education, Core Curriculum for Basic Education](https://www.oph.fi/en/education-and-qualifications/national-core-curriculum-primary-and-lower-secondary-basic-education) | 官方页面正文 |

| E09 | [Estonia Ministry of Education and Research, General Education](https://hm.ee/en/education-research-and-youth-affairs/general-education/general-education-estonia) | 官方页面介绍的 2014 框架；不是当前法律全量审计 |

| E10 | [Japan NIER, Science Education at the Stage of Compulsory Education in Japan (2021)](https://www.nier.go.jp/English/educationjapan/pdf/20210623-02.pdf) | 官方 PDF，重点为第 16–18 页 |

| E11 | [MIT Teaching + Learning Lab, Backward Design](https://tll.mit.edu/teaching-resources/course-design/backward-design/) | 官方索引内容；直接访问返回 HTTP 502 |

| E12 | [CMU Eberly Center, Learning Principles](https://www.cmu.edu/teaching/principles/learning.html) | 官方页面正文 |

| E13 | [Harvard Project Zero, Teaching for Understanding](https://pz.gse.harvard.edu/projects/teaching-for-understanding) | 官方索引内容；直接访问返回 HTTP 502 |

| E14 | [Harvard Project Zero, Visible Thinking](https://pz.gse.harvard.edu/projects/visible-thinking) | 官方索引内容 |

| E15 | [Harders & Ebersbach, No Causal Self-Explanation Effect for Factual Knowledge (2026)](https://onlinelibrary.wiley.com/doi/10.1002/acp.70174) | 原始论文摘要与相关背景；未声称完成全部方法或分析审查 |

- [GNU AGPL version 3](https://www.gnu.org/licenses/agpl-3.0.html), especially sections 2, 4–6, and 13.

- [GNU license FAQ](https://www.gnu.org/licenses/gpl-faq.html), including private modifications, aggregation, and program output.

- [GNU guidance on applying the licenses](https://www.gnu.org/licenses/gpl-howto.html).

再分发时保留版权声明、完整协议并按要求提供对应源码，维护修改日期与说明。参考 [GNU AGPL v3](https://www.gnu.org/licenses/agpl-3.0.html)、[GNU FAQ](https://www.gnu.org/licenses/gpl-faq.html) 与[应用协议说明](https://www.gnu.org/licenses/gpl-howto.html)。团队联系：[info@karmenloop.com](mailto:i

[Credential-related source omitted]

[Credential-related source omitted]

[Credential-related source omitted]

[Credential-related source omitted]

[Credential-related source omitted]

[Credential-related source omitted]

[Credential-related source omitted]

[Credential-related source omitted]

[Credential-related source omitted]

[Credential-related source omitted]

[Credential-related source omitted]

[Credential-related source omitted]

[Credential-related source omitted]

[Credential-related source omitted]

[Credential-related source omitted]

[Credential-related source omitted]

[Credential-related source omitted]

[Credential-related source omitted]

[Credential-related source omitted]

[Credential-related source omitted]

LangGraph distinguishes thread-scoped checkpoints from a cross-thread store. Its documentation explicitly warns that an in-memory checkpointer loses state on restart. [Persistence](https://docs.langchain.com/oss/python/langgraph/persistence

**Decision:** use the separation as a design reference. Do not introduce another authoritative database or a mandatory LangSmith service. **License: MIT.** [Pinned license](https://github.com/langchain-ai/langgraph/blob/3c916d750575b22c4647

The old letta-ai/letta repository now directs users to letta-ai/letta-code; its retired V1 server remains on an archive branch. [Migration notice](https://github.com/letta-ai/letta/blob/main/README.md).

Current Letta documentation describes Git-backed MemFS and background subagents that review conversations and consolidate memory. A second Agent can review proposed changes. The current harness also exposes memory, skills, prompt changes, a

**Decision:** study revisioned memory and quiet review; do not import its personhood claims, migrate user data, or install a persistent service. **License: Apache-2.0.** [Pinned license](https://github.com/letta-ai/letta-code/blob/9df436cff

DSPy optimizers use a program, an evaluation metric, and input examples. Some tune instructions or demonstrations; others fine-tune weights. Its documentation explicitly separates BootstrapFinetune and mixed optimization from prompt-only me

**Decision:** borrow explicit signatures and modular optimization, not fine-tuning, distillation, or RL. Do not adopt documentation prices or timings as PAG results. **License: MIT.** [Pinned license](https://github.com/stanfordnlp/dspy/blo

GEPA uses execution feedback to propose prompt changes and retains candidates with complementary strengths. Its adapter separates execution, feedback construction, proposal generation, and state restoration. This is a close research referen

**Decision:** use research principles and an optional evaluator adapter; independently implement the candidate lifecycle. Do not copy the optimizer or claim its research as ours. **License: MIT.** [Pinned license](https://github.com/gepa-ai

The current main repository describes Agent Canvas. The software-agent-sdk repository owns the Agent Server, tools, events, and execution; the automation repository owns schedules, webhooks, history, and dispatch. Canvas can select backends

**License: MIT for the inspected main repository.** Subrepositories and dependencies require separate review. [Pinned license](https://github.com/OpenHands/OpenHands/blob/7f6c52f8ea7187249dfe1b363a4c1efbfa9a6bc6/LICENSE).

The current main line is v2. It separates model, environment, and Agent; the README describes a linear history and independently executed shell actions. [Pinned README](https://github.com/SWE-agent/mini-swe-agent/blob/04d809ceab9df28f9adaed

**Decision:** use it as a baseline design reference, not a substitute for the actual Koder, Cursor, or Codex Agent identity. **License: MIT, in LICENSE.md.** An initial LICENSE request returned 404; the correct file was then read. [Pinned l

Inspect separates tasks, solvers, and scorers, supports resumable evaluation sets, and records evaluation logs. Its limits distinguish working time from elapsed time; the documentation notes that concurrency can make working time undercount

**Decision:** use its evaluation structure or an optional external evaluator, not its entire benchmark catalog. **License: MIT.** [Pinned license](https://github.com/UKGovernmentBEIS/inspect_ai/blob/706b4c5d6a3af83150ae3fc66ddf84b6bc37f50f/

Promptfoo documents deterministic assertions, model grading, custom metrics, trajectory checks, and latency/cost measurements. It distinguishes weighted scores from checks that must all pass. [Assertions and metrics](https://www.promptfoo.d

**Decision:** use layered criteria; do not initiate production red teaming. **License: MIT.** [Pinned license](https://github.com/promptfoo/promptfoo/blob/f2b012e3c9f9ffb7b23edf4cf88ea63febf72f5a/LICENSE).

The SDK exposes standard tools, resources, and prompts through clients and servers. The current README points to v2 documentation and Python 3.11+, so older examples are not a safe installation contract. [Pinned README](https://github.com/m

**Decision:** keep MCP optional. Preserve a file/CLI path without a third-party SDK prerequisite. **License: MIT.** [Pinned license](https://github.com/modelcontextprotocol/python-sdk/blob/d7845c073a2b9ba5ab2817ecf93d3e45ef24cd14/LICENSE).

The current README says the project is no longer maintained and recommends Microsoft Agent Framework. The repository API returned archived=false in the same review. Record that discrepancy rather than silently selecting one status. [Pinned 

GitHub's license summary is also insufficient: LICENSE-CODE is MIT while the root LICENSE contains CC-BY-4.0. [Code license](https://github.com/microsoft/autogen/blob/d07c667a3239ea4ab7af62a53b0f4ff205a1f23d/LICENSE-CODE), [root license](ht

- 核实机制：checkpointer 存同一 thread 的图状态，store 存跨 thread 的应用知识。官方文档特别指出 InMemorySaver 重启丢失；“有 checkpoint API”不等于已经持久化。[Persistence](https://docs.langchain.com/oss/python/langgraph/persistence)、[Checkpointers](https://docs.langchain.com/oss/pytho

- 建议验收：完成一次有副作用的动作，在回执写前/后分别中断；重启实际 Agent 后不得重复该副作用；同一经验在另一新会话可检索，并准确引用原 Markdown。 - 当前选择：原理参考；本轮不引入 LangGraph DB 作为第二真相源，也不增加 LangSmith 云依赖。 - 许可：MIT，已读原件。[LICENSE](https://github.com/langchain-ai/langgraph/blob/3c916d750575b22c464790e9774

- 许可：MIT，已读原件。[LICENSE](https://github.com/langchain-ai/langgraph/blob/3c916d750575b22c464790e9774494d810ccd322/LICENSE)。

- 当前项目迁移：letta-ai/letta 的 README 指向 letta-ai/letta-code；旧 V1 server 留在 archive 分支。[官方迁移说明](https://github.com/letta-ai/letta/blob/main/README.md)。

- 核实机制：MemFS 用 Git 保存可编辑记忆；dreaming 用后台 subagent 复查会话、整合经验；可再由一个 Agent 审查改动；官方描述允许修改 memory/skills/prompts/harness mods。[Memory & dreaming](https://docs.letta.com/configuration/memory)、[当前 README](https://github.com/letta-ai/letta-code/blob

- 建议验收：后台候选包含一条真实新结论、一条过时偏好、一条不具证据的推断；只合并新结论，过时偏好保留修订缘由，推断保持观察中。关闭宿主后的唤醒需另外由宿主真实执行证明。 - 当前选择：研究低干扰复盘、记忆版本、mods 差异；不移植其人格宣传，不迁移用户数据库，不默认安装常驻服务。 - 许可：Apache-2.0，已读条款与文件标识。[LICENSE](https://github.com/letta-ai/letta-code/blob/9df436cffa588f45a

- 许可：Apache-2.0，已读条款与文件标识。[LICENSE](https://github.com/letta-ai/letta-code/blob/9df436cffa588f45aadef86f34d43fb8f3008c40/LICENSE)。

- 核实机制：program + metric + input samples；可优化指令、few-shot 示例，也支持 fine-tuning。官方将 BootstrapFinetune 和 BetterTogether 与提示词优化区分，不能把所有 DSPy 优化都叫“冻结权重”。[优化器官方源文档](https://github.com/stanfordnlp/dspy/blob/c8176a8fa65e9772d9d44a33f464f4271c97b806/doc

- 建议验收：同模型同任务预算，用未优化 baseline、通用改法、领域改法、组合改法四组输出；训练/开发任务用于修改，封存任务只作最后评估，不读其答案来写候选。 - 当前选择：参考签名与分模块优化；不引入 BootstrapFinetune、权重蒸馏或 RL；不承诺框架文档中的时间/价格适用于本项目。 - 许可：MIT。[LICENSE](https://github.com/stanfordnlp/dspy/blob/c8176a8fa65e9772d9d44a33f4

- 许可：MIT。[LICENSE](https://github.com/stanfordnlp/dspy/blob/c8176a8fa65e9772d9d44a33f464f4271c97b806/LICENSE)。

- 核实机制：使用执行轨迹与文本反馈提出提示词更新，以候选集合保留不同任务上的优势，能组合互补候选；当前接口把 evaluate、reflective dataset、proposal、adapter state 分离。[论文 v2](https://arxiv.org/abs/2507.19457v2)、[官方接口](https://github.com/gepa-ai/gepa/blob/8c7b697a71977980afc7537dac0881e784eef99f/s

- 建议验收：同一任务4组的完整质量、工作时间/端到端时间、token/实际计费、回滚/副作用分别保存；若 C 破坏已有能力即拒合并，即使某个局部分数更好。 - 当前选择：原理参照与可选实验 adapter；自研候选状态机，不直接复制优化器源码或宣布发明 GEPA。 - 许可：MIT。[LICENSE](https://github.com/gepa-ai/gepa/blob/8c7b697a71977980afc7537dac0881e784eef99f/LICENSE)。

- 许可：MIT。[LICENSE](https://github.com/gepa-ai/gepa/blob/8c7b697a71977980afc7537dac0881e784eef99f/LICENSE)。

- 当前核实：主仓现在描述 Agent Canvas；实际 Agent Server/工具/事件属于 software-agent-sdk，定时与 webhook/run history 属于 automation。主仓支持选择本机、VM、容器或远端 backend，并使用 ACP 接入不同 Agent。[当前 README 与仓库边界](https://github.com/OpenHands/OpenHands/blob/7f6c52f8ea7187249dfe1b363

- 建议验收：record host automation ID + native event ID + scheduled/actual times + run ID；从真实 event 恢复同 memory root，产物读取成功才闭环。多对话共享知识、独立工作区、写入冲突可检测。 - 当前选择：借鉴控制面/执行面分离；不默认部署全套 Canvas 或连外部渠道。对方 README 标注无 sandbox 时 Agent 能读写宿主文件，本项目不能因“框架开源”自动提升权

- 许可：本次核查主仓 MIT；其他子仓和依赖要分别检查，不能由主仓许可证覆盖全部生态。[LICENSE](https://github.com/OpenHands/OpenHands/blob/7f6c52f8ea7187249dfe1b363a4c1efbfa9a6bc6/LICENSE)。

- 核实机制：当前主线 v2；模型、environment、Agent 相互分离；线性历史，shell 动作独立执行，轨迹可以直接审查。[官方 README](https://github.com/SWE-agent/mini-swe-agent/blob/04d809ceab9df28f9adaed044884180159172930/README.md)。

- 建议验收：用同一真实缺陷修复任务与同一解释/教学任务做最小 baseline；记录总调用和实际完成质量，避免只比较 Agent 核心代码行数。 - 当前选择：比较对象与简化思想，不替换 Koder/Cursor/Codex 的实际运行身份。 - 许可：MIT，文件实际是 LICENSE.md（最初 LICENSE 返回404，随后定位正确文件）。[LICENSE.md](https://github.com/SWE-agent/mini-swe-agent/blob/04

- 许可：MIT，文件实际是 LICENSE.md（最初 LICENSE 返回404，随后定位正确文件）。[LICENSE.md](https://github.com/SWE-agent/mini-swe-agent/blob/04d809ceab9df28f9adaed044884180159172930/LICENSE.md)。

- 核实机制：独立 task/solver/scorer；eval-set 支持恢复未完成的评估；日志保存执行信息；限制可区分端到端时间与工作时间，并记录 token/费用。官方指出并行等待会使 working time 低估，不能用它替代 wall time。[日志](https://inspect.aisi.org.uk/eval-logs.html)、[eval-set](https://inspect.aisi.org.uk/eval-sets.html)、[限制](

- 建议验收：中断后的恢复不重复成功样本；失败/取消/超时/未知各自统计；费用未知写 null + 原因，不写0。任何评分改动保留版本及历史，不覆盖以前结果。 - 当前选择：原理与可选 external evaluator；本轮不安装200多个评估、不因通过某数据集就声称通用精通。 - 许可：MIT。[LICENSE](https://github.com/UKGovernmentBEIS/inspect_ai/blob/706b4c5d6a3af83150ae3fc66dd

- 许可：MIT。[LICENSE](https://github.com/UKGovernmentBEIS/inspect_ai/blob/706b4c5d6a3af83150ae3fc66ddf84b6bc37f50f/LICENSE)。

- 核实机制：确定性断言、模型评分、自定义指标、轨迹检查、latency/cost 指标；文档区分加权得分与必须每项通过，支持规则分组。[Assertions & metrics](https://www.promptfoo.dev/docs/configuration/expected-outputs/)。

- 建议验收：加入格式全对但内容错误、非常简短但关键约束丢失、旧缓存回执、工具调用成功但产物错误等受控反例；盲评方法固定，随机对调候选显示顺序。 - 当前选择：参考分层判据；不引入生产攻击扫描，不拿通用 red-team 标签作为任务质量。 - 许可：MIT。[LICENSE](https://github.com/promptfoo/promptfoo/blob/f2b012e3c9f9ffb7b23edf4cf88ea63febf72f5a/LICENSE)。

- 许可：MIT。[LICENSE](https://github.com/promptfoo/promptfoo/blob/f2b012e3c9f9ffb7b23edf4cf88ea63febf72f5a/LICENSE)。

- 核实机制：标准 tools/resources/prompts，server/client 与 stdio/Streamable HTTP 等 transport；当前 README 是 Python3.11+ 的 v2 文档路径，不能沿用旧 API 片段。[README](https://github.com/modelcontextprotocol/python-sdk/blob/d7845c073a2b9ba5ab2817ecf93d3e45ef24cd14/REA

- 建议验收：严格路径范围、大小上限、revision 对比写；未知工具拒绝，读工具不能触发写；外部网页和工具描述不能覆盖用户约束。跨宿主读取同一 Markdown 字节，并在恢复后写回 revision。 - 当前选择：记录互操作契约，支持可选 MCP adapter；先保留无第三方依赖的 CLI/file 路径，不让安装整个SDK变成使用Skill前提。 - 许可：MIT。[LICENSE](https://github.com/modelcontextprotocol/

- 许可：MIT。[LICENSE](https://github.com/modelcontextprotocol/python-sdk/blob/d7845c073a2b9ba5ab2817ecf93d3e45ef24cd14/LICENSE)。

- 当前 README 明确 no longer maintained/archived 并建议新项目选择 Microsoft Agent Framework；同次 GitHub repository API 的 archived=false，二者状态不一致。应写“README 宣布停止维护、API归档标志未同步”，而不是擅自抹平冲突。[README](https://github.com/microsoft/autogen/blob/d07c667a3239ea4ab7af

- 许可不能只看GitHub summary：代码 LICENSE-CODE 为MIT，根LICENSE为CC-BY-4.0文本。[代码许可](https://github.com/microsoft/autogen/blob/d07c667a3239ea4ab7af62a53b0f4ff205a1f23d/LICENSE-CODE)、[根许可](https://github.com/microsoft/autogen/blob/d07c667a3239ea4ab7af62a

Mechanism reference: the [official Harness² repository](https://github.com/google-research/harness2) describes Propose–Probe–Compose and persistent improvement experience with frozen model weights. This skill implements its own host-driven 

"result": "PASS",     "run_id": 39480710426,     "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/39480710426",

"run_id": 39480710426,     "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/39480710426",     "commit": "43e6f6a8323a1aa8ee99cfcb3c96f7ae271014ca",

"url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/39480710426",

"name": "check (ubuntu-24.04)",         "job_id": 118828210523,         "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/39480710426/job/118828210523",

"job_id": 118828210523,         "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/39480710426/job/118828210523",         "conclusion": "success",

"url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/39480710426/job/118828210523",

"name": "check (macos-26-intel)",         "job_id": 118828210858,         "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/39480710426/job/118828210858",

"job_id": 118828210858,         "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/39480710426/job/118828210858",         "conclusion": "success",

"url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/39480710426/job/118828210858",

"name": "check (windows-2025)",         "job_id": 118828211169,         "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/39480710426/job/118828211169",

"job_id": 118828211169,         "url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/39480710426/job/118828211169",         "conclusion": "success",

"url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/39480710426/job/118828211169",

Verified on October 21, 2026, at 03:09 Asia/Shanghai. The [completed workflow](https://github.com/KarmenLoop/personal-agent-growth/actions/runs/39480710426) tested commit `43e6f6a8323a1aa8ee99cfcb3c96f7ae271014ca`. The [machine-readable rec

| Environment | Python | Utility tests | Package checks | Job | |---|---|---|---|---| | Ubuntu 24.04, x64 | 3.11.17 | 13 passed | 21 passed | [Linux log](https://github.com/KarmenLoop/personal-agent-growth/actions/runs/39480710426/job/11882

|---|---|---|---|---| | Ubuntu 24.04, x64 | 3.11.17 | 13 passed | 21 passed | [Linux log](https://github.com/KarmenLoop/personal-agent-growth/actions/runs/39480710426/job/118828210523) | | macOS 26.6.1, Intel x64 | 3.11.9 | 13 passed | 21 p

| Ubuntu 24.04, x64 | 3.11.17 | 13 passed | 21 passed | [Linux log](https://github.com/KarmenLoop/personal-agent-growth/actions/runs/39480710426/job/118828210523) |

| macOS 26.6.1, Intel x64 | 3.11.9 | 13 passed | 21 passed | [macOS log](https://github.com/KarmenLoop/personal-agent-growth/actions/runs/39480710426/job/118828210858) |

| Windows Server 2025, x64 | 3.11.9 | 13 passed | 21 passed | [Windows log](https://github.com/KarmenLoop/personal-agent-growth/actions/runs/39480710426/job/118828211169) |

The [initial main-branch run](https://github.com/KarmenLoop/personal-agent-growth/actions/runs/39468002952) and [release-tag run](https://github.com/KarmenLoop/personal-agent-growth/actions/runs/39468117247) could not acquire a `macos-lates

The [release download](https://github.com/KarmenLoop/personal-agent-growth/releases/tag/v0.3.2) is unchanged. The full AGPL-3.0-only license and copyright notice remain in both the repository and the package.

| Runner or shell | Result | Actual run | |---|---|---| | Ubuntu 24.04, macOS 26 Intel, Windows Server 2025 | Each passed 48 tests, core build/123 checks, and publication checks | [Corrected implementation run](https://github.com/KarmenLoop

|---|---|---| | Ubuntu 24.04, macOS 26 Intel, Windows Server 2025 | Each passed 48 tests, core build/123 checks, and publication checks | [Corrected implementation run](https://github.com/KarmenLoop/personal-agent-growth/actions/runs/403872

| Ubuntu 24.04, macOS 26 Intel, Windows Server 2025 | Each passed 48 tests, core build/123 checks, and publication checks | [Corrected implementation run](https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40387260224) |

| Windows PowerShell 5.1.26100.33648 and PowerShell 7.6.7 | Both passed default D: and custom C: Unicode/space paths | [Guide and shared-file run](https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40387005475) |

The [first CI run](https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40387006023) had one Windows failure: a test compared a short `RUNNER~1` path with its resolved long spelling. It now verifies the actual indexed directory 

Final public assets and post-upload download receipts are available on [v0.4.1](https://github.com/KarmenLoop/personal-agent-growth/releases/tag/v0.4.1). `RELEASE-VERIFICATION.json` records the published commit, downloaded hashes, package c

| 系统或 Shell | 结果 | 实际运行 | |---|---|---| | Ubuntu 24.04、macOS 26 Intel、Windows Server 2025 | 各自 48 项测试、核心构建/123 项检查、文档检查全部通过 | [修正后的实现检查](https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40387260224) |

|---|---|---| | Ubuntu 24.04、macOS 26 Intel、Windows Server 2025 | 各自 48 项测试、核心构建/123 项检查、文档检查全部通过 | [修正后的实现检查](https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40387260224) | | Windows PowerShell 5.1.26100.33648、PowerShell 7

| Ubuntu 24.04、macOS 26 Intel、Windows Server 2025 | 各自 48 项测试、核心构建/123 项检查、文档检查全部通过 | [修正后的实现检查](https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40387260224) |

| Windows PowerShell 5.1.26100.33648、PowerShell 7.6.7 | 默认 D 盘及自定义 C 盘中文空格路径全部通过 | [安装与共享文件检查](https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40387005475) |

[首轮 CI](https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40387006023) 有一项 Windows 失败：测试把 RUNNER~1 短路径与解析后的完整路径按文字比较。现改为 samefile 核实真实索引目录，快照与篡改检查全部保留。旧失败没有删除；这是测试可移植性修正，不代表经验加载器返回了错误目录。

正式产物及上传后下载回执见 [v0.4.1](https://github.com/KarmenLoop/personal-agent-growth/releases/tag/v0.4.1)。RELEASE-VERIFICATION.json 记录发布提交、下载哈希、包检查及真实网络安装结果；该文件缺失时不能认为这些发布后检查已完成。它位于 ZIP 外，避免压缩包哈希自引用。

The [0.4.2 release](https://github.com/KarmenLoop/personal-agent-growth/releases/tag/v0.4.2) includes both the core Skill and the complete bilingual offline bundle. `RELEASE-VERIFICATION.json` is kept outside the ZIP and binds the published

The first hosted Windows installation run stopped before installation because its transport adapter still named 0.4.1. The adapter now reads the current publication version, validates the local manifest, and derives the download directory. 

Hosted verification subsequently passed: [50 tests and package checks on Ubuntu, macOS Intel, and Windows](https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40395595837), plus [PowerShell 5.1 and 7 installation, shared-file r

[0.4.2 发布页](https://github.com/KarmenLoop/personal-agent-growth/releases/tag/v0.4.2)提供核心 Skill 与完整双语离线包。ZIP 外单独的 `RELEASE-VERIFICATION.json` 绑定公开提交、下载附件哈希与发布后真实安装，避免归档校验值自引用。[0.4.1 验收](release-0.4.1-acceptance.zh-CN.md)及旧失败按原日期保留。

首轮 Windows 安装 CI 在安装前停止，原因是本地传输适配仍固定 0.4.1。现读取当前发布版本、核对本地 manifest 并推导下载目录；新增两项检查，覆盖未来版本、错误清单和语言不一致。[失败 CI 保留](https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40394859606)。九个内嵌 Mermaid 图均已解析通过，见[图示回执](evidence/release-0.4.2

后续宿主验证通过：[Ubuntu、macOS Intel、Windows 各 50 项测试及包检查](https://github.com/KarmenLoop/personal-agent-growth/actions/runs/40395595837)，以及 [PowerShell 5.1/7 安装、共享文件修订 1→2→3 和受保护卸载](https://github.com/KarmenLoop/personal-agent-growth/actions/runs/4

The installable core includes **68 files**, including the helper and both runtime/tutorial languages. Source, tests, research, diagrams, history, attribution and license are included in the complete delivery. Local build/install/relocation/

核心包 **68文件**，包含新脚本、双语运行协议与教程；完整包另有源码、测试、研究、图源、历史、团队与协议。本地打包、安装、搬迁和卸载保留记忆见[打包回执](evidence/capability-learning-20261103/package-checks.json)。最终GitHub CI与公开下载读回另绑定发布页的 [RELEASE-VERIFICATION.json](https://github.com/KarmenLoop/personal-agent-gr

"status": "PASS",   "run_id": 39485618391,   "run_url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/39485618391",

"run_id": 39485618391,   "run_url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/39485618391",   "tested_commit": "89c3bddefc95b788e320bb117c0e8f4324a9a566",

"run_url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/39485618391",

"os": "Microsoft Windows Server 2025 Datacenter",       "python": "3.11.9 (tags/v3.11.9:de54cf5, Apr  2 2024, 10:12:12) [MSC v.1938 64 bit (AMD64)]",       "job_url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/3948561

"python": "3.11.9 (tags/v3.11.9:de54cf5, Apr  2 2024, 10:12:12) [MSC v.1938 64 bit (AMD64)]",       "job_url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/39485618391/job/118845129525",       "guide_sha256": {

"job_url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/39485618391/job/118845129525",

"os": "Microsoft Windows Server 2025 Datacenter",       "python": "3.11.9 (tags/v3.11.9:de54cf5, Apr  2 2024, 10:12:12) [MSC v.1938 64 bit (AMD64)]",       "job_url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/3948561

"python": "3.11.9 (tags/v3.11.9:de54cf5, Apr  2 2024, 10:12:12) [MSC v.1938 64 bit (AMD64)]",       "job_url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/39485618391/job/118845129868",       "guide_sha256": {

"job_url": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/39485618391/job/118845129868",

}   ],   "initial_failed_run": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/39485238116",

],   "initial_failed_run": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/39485238116",   "fixes": [

"initial_failed_run": "https://github.com/KarmenLoop/personal-agent-growth/actions/runs/39485238116",

**PASS** on real GitHub-hosted Windows Server 2025 x64 runners. The [completed run](https://github.com/KarmenLoop/personal-agent-growth/actions/runs/39485618391) tested commit `89c3bddefc95b788e320bb117c0e8f4324a9a566`.

The [first run](https://github.com/KarmenLoop/personal-agent-growth/actions/runs/39485238116) remains available and is included in the evidence archive.

**Windows PowerShell module lookup.** Launching PowerShell 5.1 through Python under PowerShell 7 inherited the wrong module search paths. `Get-FileHash` was unavailable before extraction. The test runner now removes `PSModulePath` only from

**Captured Unicode output.** The C: Chinese-path review initially received unusable output. That first wrapper record retained the JSON parsing failure but did not export the nested verifier's raw error. The follow-up retained a direct repr

- [Machine-readable results](windows-install-results.json), including exact shell versions, cases, guide hashes, job URLs, and receipt hashes. - [Full evidence ZIP](https://github.com/KarmenLoop/personal-agent-growth/releases/download/v0.3.

- [Full evidence ZIP](https://github.com/KarmenLoop/personal-agent-growth/releases/download/v0.3.2/windows-install-evidence-39485618391.zip), containing the first failure and successful follow-up commands and outputs.

Copyright (C) 2007 Free Software Foundation, Inc. <https://fsf.org/>

You should have received a copy of the GNU Affero General Public License     along with this program.  If not, see <https://www.gnu.org/licenses/>.

along with this program.  If not, see <https://www.gnu.org/licenses/>.

if any, to sign a "copyright disclaimer" for the program, if necessary. For more information on this, and how to apply and follow the GNU AGPL, see <https://www.gnu.org/licenses/>.

<https://www.gnu.org/licenses/>.

Copyright (C) 2026 KARMENLOOP Official Team Contact: info@karmenloop.com Project: https://github.com/KarmenLoop/personal-agent-growth

Contact: info@karmenloop.com Project: https://github.com/KarmenLoop/personal-agent-growth SPDX-License-Identifier: AGPL-3.0-only

Project: https://github.com/KarmenLoop/personal-agent-growth

You should have received a copy of the GNU Affero General Public License along with this program. If not, see https://www.gnu.org/licenses/.

along with this program. If not, see https://www.gnu.org/licenses/.

Executable file

#!/usr/bin/env python3

Copyright (C) 2007 Free Software Foundation, Inc. <https://fsf.org/>

You should have received a copy of the GNU Affero General Public License     along with this program.  If not, see <https://www.gnu.org/licenses/>.

along with this program.  If not, see <https://www.gnu.org/licenses/>.

if any, to sign a "copyright disclaimer" for the program, if necessary. For more information on this, and how to apply and follow the GNU AGPL, see <https://www.gnu.org/licenses/>.

<https://www.gnu.org/licenses/>.

Copyright (C) 2026 KARMENLOOP Official Team Contact: info@karmenloop.com Project: https://github.com/KarmenLoop/personal-agent-growth

Contact: info@karmenloop.com Project: https://github.com/KarmenLoop/personal-agent-growth SPDX-License-Identifier: AGPL-3.0-only

Project: https://github.com/KarmenLoop/personal-agent-growth

You should have received a copy of the GNU Affero General Public License along with this program. If not, see https://www.gnu.org/licenses/.

along with this program. If not, see https://www.gnu.org/licenses/.

机制参考：[Harness² 官方仓库](https://github.com/google-research/harness2) 与 [原论文](https://arxiv.org/abs/2610.14224)。此 skill 独立实现其 Propose–Probe–Compose 和双层经验反馈思路；额外使用固定质量约束、组合复验、修订检查及回滚。没有集成上游运行器，也不沿用其基准成绩作为本项目验收。

Executable file

Executable file

#!/usr/bin/env python3

Executable file

#!/usr/bin/env python3

Research indicates that human feedback may reward sycophancy under some conditions, not that all RLHF necessarily does so. See [Towards Understanding Sycophancy in Language Models](https://arxiv.org/abs/2310.13548) and OpenAI's [specific in

研究说明人类反馈在一些条件下可能奖励迎合；不意味着所有 RLHF 都必然如此。参阅 [Towards Understanding Sycophancy in Language Models](https://arxiv.org/abs/2310.13548) 与 OpenAI 的[具体事故分析](https://openai.com/index/expanding-on-sycophancy/)（查阅：2026-11-03）。本协议由本项目编写，未引入第三方训练或运行框架。

Mechanism reference: [official Harness² repository](https://github.com/google-research/harness2) and [paper](https://arxiv.org/abs/2610.14224). This project independently implements the idea with fixed quality gates, composition retesting, 

机制参考：[Harness² 官方仓库](https://github.com/google-research/harness2) 与 [原论文](https://arxiv.org/abs/2610.14224)。此 skill 独立实现其 Propose–Probe–Compose 和双层经验反馈思路；额外使用固定质量约束、组合复验、修订检查及回滚。没有集成上游运行器，也不沿用其基准成绩作为本项目验收。

created: 2026-10-27 updated: 2026-10-27 sources: [https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.3/docs/evidence/koder-skip-audit-20261027.json, https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.2/docs/evidence/ko

updated: 2026-10-27 sources: [https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.3/docs/evidence/koder-skip-audit-20261027.json, https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.2/docs/evidence/koder-regression-repai

sources: [https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.3/docs/evidence/koder-skip-audit-20261027.json, https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.2/docs/evidence/koder-regression-repair-20261027.json, htt

created: 2026-10-27 updated: 2026-10-27 sources: [https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.3/docs/evidence/koder-skip-audit-20261027.json, https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.2/docs/evidence/ko

updated: 2026-10-27 sources: [https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.3/docs/evidence/koder-skip-audit-20261027.json, https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.2/docs/evidence/koder-regression-repai

sources: [https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.3/docs/evidence/koder-skip-audit-20261027.json, https://github.com/KarmenLoop/personal-agent-growth/blob/v0.4.2/docs/evidence/koder-regression-repair-20261027.json, htt

created: 2026-10-27 updated: 2026-10-28 sources: [https://learn.microsoft.com/en-us/windows-server/administration/windows-commands/icacls]

updated: 2026-10-28 sources: [https://learn.microsoft.com/en-us/windows-server/administration/windows-commands/icacls] ---

sources: [https://learn.microsoft.com/en-us/windows-server/administration/windows-commands/icacls]

created: 2026-10-27 updated: 2026-10-28 sources: [https://learn.microsoft.com/en-us/windows-server/administration/windows-commands/icacls]

updated: 2026-10-28 sources: [https://learn.microsoft.com/en-us/windows-server/administration/windows-commands/icacls] ---

sources: [https://learn.microsoft.com/en-us/windows-server/administration/windows-commands/icacls]

#!/usr/bin/env python3

#!/usr/bin/env python3

#!/usr/bin/env python3

#!/usr/bin/env python3

#!/usr/bin/env python3

#!/usr/bin/env python3

#!/usr/bin/env python3

#!/usr/bin/env python3

Executable file

#!/usr/bin/env python3

[Credential-related source omitted]

#!/usr/bin/env python3

#!/usr/bin/env python3

Executable file

Executable file

Executable file

malformed = [             {**original, "path": "../outside.md"},             {**original, "path": "https://example.invalid/experience.md"},

{**original, "path": "../outside.md"},             {**original, "path": "https://example.invalid/experience.md"},             {**original, "head_sha256": "0" * 64},

{**original, "path": "https://example.invalid/experience.md"},

Executable file

Executable file

Executable file

Executable file

Executable file

Executable file

Executable file

Executable file

Executable file

Executable file

Executable file

text = archive.read("personal-agent-growth/README.md").decode("utf-8")             version = json.loads(self.manifest.read_text())["declared_version"]             self.assertIn(f"https://github.com/KarmenLoop/personal-agent-growth/blob/v{ve

version = json.loads(self.manifest.read_text())["declared_version"]             self.assertIn(f"https://github.com/KarmenLoop/personal-agent-growth/blob/v{version}/docs/architecture.en.md", text)             self.assertNotIn("](docs/archite

self.assertIn(f"https://github.com/KarmenLoop/personal-agent-growth/blob/v{version}/docs/architecture.en.md", text)

Executable file

#!/usr/bin/env python3

#!/usr/bin/env python3

if relative.startswith("../") or not (ROOT / relative).exists():             raise ValueError("Unresolved package document link: " + name + " -> " + target)         url = "https://github.com/KarmenLoop/personal-agent-growth/blob/v" + versio

raise ValueError("Unresolved package document link: " + name + " -> " + target)         url = "https://github.com/KarmenLoop/personal-agent-growth/blob/v" + version + "/" + quote(relative)         return "](" + url + ("#" + fragment if sep 

url = "https://github.com/KarmenLoop/personal-agent-growth/blob/v" + version + "/" + quote(relative)

#!/usr/bin/env python3

#!/usr/bin/env python3

[Credential-related source omitted]

[Credential-related source omitted]

[Credential-related source omitted]

#!/usr/bin/env python3

#!/usr/bin/env python3

Executable file

#!/usr/bin/env python3

Executable file

browser = await chromium.launch({ headless: true, ...(args.includes('--browser') ? { executablePath: option('--browser') } : {}) });   const page = await browser.newPage({ viewport: { width: 1600, height: 1000 } });   await page.goto(`http:

const page = await browser.newPage({ viewport: { width: 1600, height: 1000 } });   await page.goto(`http://127.0.0.1:${server.address().port}/`);   await page.evaluate(async () => {

await page.goto(`http://127.0.0.1:${server.address().port}/`);